On 12/18/2023 2:13 AM, Bjorn Andersson wrote: > On Sat, Nov 25, 2023 at 12:20:59AM +0530, Vignesh Viswanathan wrote: >> q6v5 fatal and watchdog IRQ handlers always retrieves the crash reason >> information from SMEM global partition (QCOM_SMEM_HOST_ANY). >> >> For some targets like IPQ9574 and IPQ5332, crash reason information is >> present in target specific partition due to which the crash reason is >> not printed in the current implementation. >> >> Add support to pass crash_reason_smem_id along with crash_reason item >> number in qcom_q6v5_init call and use the same to get the crash >> information from SMEM in fatal and watchdog IRQ handlers. >> >> This patch depends on [1] which adds support for IPQ9574 and IPQ5332 >> remoteproc q5v5_mpd driver. > > This is solely here to ensure things are applied in appropriate order, > there's no benefit in documenting it in the eternal git history. So > please move this comment below the "---" line. > Ack. Will move this below in V2. >> >> [1]: https://lore.kernel.org/all/20231110091939.3025413-1-quic_mmanikan@xxxxxxxxxxx/ >> >> Signed-off-by: Vignesh Viswanathan <quic_viswanat@xxxxxxxxxxx> >> --- >> drivers/remoteproc/qcom_q6v5.c | 10 +++++++--- >> drivers/remoteproc/qcom_q6v5.h | 4 +++- >> drivers/remoteproc/qcom_q6v5_adsp.c | 3 ++- >> drivers/remoteproc/qcom_q6v5_mpd.c | 2 +- >> drivers/remoteproc/qcom_q6v5_mss.c | 5 +++-- >> drivers/remoteproc/qcom_q6v5_pas.c | 3 ++- >> drivers/remoteproc/qcom_q6v5_wcss.c | 4 +++- >> 7 files changed, 21 insertions(+), 10 deletions(-) >> >> diff --git a/drivers/remoteproc/qcom_q6v5.c b/drivers/remoteproc/qcom_q6v5.c >> index 0e32f13c196d..072e41730110 100644 >> --- a/drivers/remoteproc/qcom_q6v5.c >> +++ b/drivers/remoteproc/qcom_q6v5.c >> @@ -100,7 +100,8 @@ static irqreturn_t q6v5_wdog_interrupt(int irq, void *data) >> return IRQ_HANDLED; >> } >> >> - msg = qcom_smem_get(QCOM_SMEM_HOST_ANY, q6v5->crash_reason, &len); >> + msg = qcom_smem_get(q6v5->crash_reason_smem_id, q6v5->crash_reason, >> + &len); > > No need to break lines that are just slightly over 80 characters... Ack. > >> if (!IS_ERR(msg) && len > 0 && msg[0]) >> dev_err(q6v5->dev, "watchdog received: %s\n", msg); >> else >> @@ -121,7 +122,8 @@ irqreturn_t q6v5_fatal_interrupt(int irq, void *data) >> if (!q6v5->running) >> return IRQ_HANDLED; >> >> - msg = qcom_smem_get(QCOM_SMEM_HOST_ANY, q6v5->crash_reason, &len); >> + msg = qcom_smem_get(q6v5->crash_reason_smem_id, q6v5->crash_reason, >> + &len); >> if (!IS_ERR(msg) && len > 0 && msg[0]) >> dev_err(q6v5->dev, "fatal error received: %s\n", msg); >> else >> @@ -279,7 +281,8 @@ EXPORT_SYMBOL_GPL(qcom_q6v5_panic); >> * Return: 0 on success, negative errno on failure >> */ >> int qcom_q6v5_init(struct qcom_q6v5 *q6v5, struct platform_device *pdev, >> - struct rproc *rproc, int crash_reason, const char *load_state, >> + struct rproc *rproc, int crash_reason, >> + int crash_reason_smem_id, const char *load_state, >> void (*handover)(struct qcom_q6v5 *q6v5)) >> { >> int ret; >> @@ -287,6 +290,7 @@ int qcom_q6v5_init(struct qcom_q6v5 *q6v5, struct platform_device *pdev, >> q6v5->rproc = rproc; >> q6v5->dev = &pdev->dev; >> q6v5->crash_reason = crash_reason; >> + q6v5->crash_reason_smem_id = crash_reason_smem_id; >> q6v5->handover = handover; >> >> init_completion(&q6v5->start_done); >> diff --git a/drivers/remoteproc/qcom_q6v5.h b/drivers/remoteproc/qcom_q6v5.h >> index 4e1bb1a68284..21cd879e6e1e 100644 >> --- a/drivers/remoteproc/qcom_q6v5.h >> +++ b/drivers/remoteproc/qcom_q6v5.h >> @@ -41,6 +41,7 @@ struct qcom_q6v5 { >> struct completion spawn_done; >> >> int crash_reason; >> + int crash_reason_smem_id; > > While this is called "smem_id" in some places, you refer to it as SMEM > partition in the commit message - and that's much less confusing. > > So please rename this. > Sure, partition makes more sense in this context. Will rename. >> >> bool running; >> >> @@ -49,7 +50,8 @@ struct qcom_q6v5 { >> }; >> >> int qcom_q6v5_init(struct qcom_q6v5 *q6v5, struct platform_device *pdev, >> - struct rproc *rproc, int crash_reason, const char *load_state, >> + struct rproc *rproc, int crash_reason, >> + int crash_reason_smem_id, const char *load_state, > > To me it would be more natural if the most significant specifier was > given before the least significant specifier. Please swap the partition > and item parameters... > > Regards, > Bjorn Ack. Thanks, Vignesh > >> void (*handover)(struct qcom_q6v5 *q6v5)); >> void qcom_q6v5_deinit(struct qcom_q6v5 *q6v5); >> >> diff --git a/drivers/remoteproc/qcom_q6v5_adsp.c b/drivers/remoteproc/qcom_q6v5_adsp.c >> index 6c67514cc493..30d91205f199 100644 >> --- a/drivers/remoteproc/qcom_q6v5_adsp.c >> +++ b/drivers/remoteproc/qcom_q6v5_adsp.c >> @@ -732,7 +732,8 @@ static int adsp_probe(struct platform_device *pdev) >> goto disable_pm; >> >> ret = qcom_q6v5_init(&adsp->q6v5, pdev, rproc, desc->crash_reason_smem, >> - desc->load_state, qcom_adsp_pil_handover); >> + QCOM_SMEM_HOST_ANY, desc->load_state, >> + qcom_adsp_pil_handover); >> if (ret) >> goto disable_pm; >> >> diff --git a/drivers/remoteproc/qcom_q6v5_mpd.c b/drivers/remoteproc/qcom_q6v5_mpd.c >> index b133285888c7..839f6a15b88d 100644 >> --- a/drivers/remoteproc/qcom_q6v5_mpd.c >> +++ b/drivers/remoteproc/qcom_q6v5_mpd.c >> @@ -726,7 +726,7 @@ static int q6_wcss_probe(struct platform_device *pdev) >> goto free_rproc; >> >> ret = qcom_q6v5_init(&wcss->q6, pdev, rproc, >> - WCSS_CRASH_REASON, NULL, NULL); >> + WCSS_CRASH_REASON, WCSS_SMEM_HOST, NULL, NULL); >> if (ret) >> goto free_rproc; >> >> diff --git a/drivers/remoteproc/qcom_q6v5_mss.c b/drivers/remoteproc/qcom_q6v5_mss.c >> index 394b2c1cb5e2..45ecb87d73ef 100644 >> --- a/drivers/remoteproc/qcom_q6v5_mss.c >> +++ b/drivers/remoteproc/qcom_q6v5_mss.c >> @@ -26,6 +26,7 @@ >> #include <linux/remoteproc.h> >> #include <linux/reset.h> >> #include <linux/soc/qcom/mdt_loader.h> >> +#include <linux/soc/qcom/smem.h> >> #include <linux/iopoll.h> >> #include <linux/slab.h> >> >> @@ -2093,8 +2094,8 @@ static int q6v5_probe(struct platform_device *pdev) >> qproc->need_mem_protection = desc->need_mem_protection; >> qproc->has_mba_logs = desc->has_mba_logs; >> >> - ret = qcom_q6v5_init(&qproc->q6v5, pdev, rproc, MPSS_CRASH_REASON_SMEM, "modem", >> - qcom_msa_handover); >> + ret = qcom_q6v5_init(&qproc->q6v5, pdev, rproc, MPSS_CRASH_REASON_SMEM, >> + QCOM_SMEM_HOST_ANY, "modem", qcom_msa_handover); >> if (ret) >> goto detach_proxy_pds; >> >> diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c >> index 913a5d2068e8..6a29b6ab181f 100644 >> --- a/drivers/remoteproc/qcom_q6v5_pas.c >> +++ b/drivers/remoteproc/qcom_q6v5_pas.c >> @@ -728,7 +728,8 @@ static int adsp_probe(struct platform_device *pdev) >> goto free_rproc; >> adsp->proxy_pd_count = ret; >> >> - ret = qcom_q6v5_init(&adsp->q6v5, pdev, rproc, desc->crash_reason_smem, desc->load_state, >> + ret = qcom_q6v5_init(&adsp->q6v5, pdev, rproc, desc->crash_reason_smem, >> + QCOM_SMEM_HOST_ANY, desc->load_state, >> qcom_pas_handover); >> if (ret) >> goto detach_proxy_pds; >> diff --git a/drivers/remoteproc/qcom_q6v5_wcss.c b/drivers/remoteproc/qcom_q6v5_wcss.c >> index cff1fa07d1de..077ecfa2cf86 100644 >> --- a/drivers/remoteproc/qcom_q6v5_wcss.c >> +++ b/drivers/remoteproc/qcom_q6v5_wcss.c >> @@ -18,6 +18,7 @@ >> #include <linux/regulator/consumer.h> >> #include <linux/reset.h> >> #include <linux/soc/qcom/mdt_loader.h> >> +#include <linux/soc/qcom/smem.h> >> #include "qcom_common.h" >> #include "qcom_pil_info.h" >> #include "qcom_q6v5.h" >> @@ -1047,7 +1048,8 @@ static int q6v5_wcss_probe(struct platform_device *pdev) >> if (ret) >> goto free_rproc; >> >> - ret = qcom_q6v5_init(&wcss->q6v5, pdev, rproc, desc->crash_reason_smem, NULL, NULL); >> + ret = qcom_q6v5_init(&wcss->q6v5, pdev, rproc, desc->crash_reason_smem, >> + QCOM_SMEM_HOST_ANY, NULL, NULL); >> if (ret) >> goto free_rproc; >> >> -- >> 2.41.0 >>
